It all became part of a massive Japanese renewal project: down came the gray office blocks where the Japanese had toiled during the tough “catching-up” years, and down came the rickety wooden houses of the nation’s “salarymen.” In their place were erected contemporary office blocks and stylish homes that expressed the Japanese pride in their wealth. Banks had found a ready market.
Japan’s most intense lending growth happened from 1988 to 1990, with nominal loans growing an average of 13.8 percent per year. During this Japanese economic miracle, CRE lending skyrocketed, and new office buildings, apartments, hotels, and shopping malls soon dominated the landscape. Japan’s largest “blue chip” corporations had gained access to corporate debt markets and were no longer as dependent on banks. This meant that Japan’s bank lenders were forced to diversify and go downmarket, so they aggressively chased small- to medium-sized businesses and real estate lending. Banks lent to these smaller domestic businesses and to consumers through mortgages and a category called “loans for other purposes” that were often secured by real estate. Japan’s city banks made almost 75 percent of their loans to small businesses, many of which were backed by real estate, though not categorized or disclosed that way. In addition, a particularly risky form of bank lending known as corporate overdrafts surged, growing “from 2 percent to 19 percent (or ¥41.6 trillion) of total outstanding loans between March 1981 and October 1991.”8
These kinds of loans were riskier than blue-chip corporate loans—and lending standards were increasingly compromised. The Bank of Japan would later find that lenders’ risk departments, the very places entrusted to inject caution into lending practices, had instead become dominated by those seeking to promote loans and who overoptimistically valued collateral.
The deregulation of interest rates that banks could pay their customers began largely in 1985 and was fully implemented by April 1991. Banks could pay more interest to customers and thus attract more deposits, an important adjunct to meteoric loan growth. The United States had deregulated its financial system over the course of several decades, from the 1970s to the 1990s. Post–Plaza Accord, it pressured the Japanese to deregulate in a matter of a few years. This was another fateful factor. In the ministries, there was no institutional experience or preparation for such swift financial deregulation. Additionally, lending banks had little understanding of the CRE market. It all increased the odds of an unfavorable outcome for the lending industry.
Japan’s lending frenzy extended down into households. Credit cards accounts tripled from 1983 to 1990, to 166 million. Nonmortgage household debt increased from 1980 to 1990, from ¥16 trillion to ¥49 trillion. From 1984 to 1990, the average size of a bank mortgage doubled from ¥5.6 million to ¥11 million.9 By 1991, Japan, a nation that until recently had no significant consumer-driven debt to speak of, saw its per-capita consumer debt grow to $3,426, an amount commensurate with that of the United States.
Japan’s lending frenzy extended abroad. In highly publicized transactions that alarmed many Americans, Hiro Real Estate purchased New York’s Socony-Mobil Building in 1987; the next year, the Bank of Tokyo bought Union Bank; in 1989, Mitsubishi Real Estate bought Rockefeller Center; and, in 1990, a Japanese firm called Cosmo World bought the legendary American golf course Pebble Beach. But “trophy” property acquisitions such as these often signal a market at the top and thus herald a tumble. And so it was here. By 1990, Japan’s banks owned an astonishing 12.4 percent of all American banking assets. In 1991, Japan’s banks accounted for 24 percent of all California banking assets, including 20 percent of all property loans and 35 percent of all commercial and industrial loans.10
But the 1980s lending frenzy wasn’t just bank based, which is typical of a financial crisis. Leasing companies, consumer-finance companies, mortgage companies, and other nonbank lenders were making significant property-backed loans, and their funding often came from banks. Bank loans to fund nonbank lenders reached some ¥137 trillion by 1991, much of which eventually went bad. During this period, Japan’s 30,000 nonbanks were largely unregulated, so the composition of their portfolios was largely unknown. This opacity “made America’s savings and loans regulators look almost zealous by comparison.”11
A 1990 finance ministry study concluded that about 60 percent of all nonbank loans had gone to the property and construction sectors compared with 25 percent at banks. Total nonbank lending would reach ¥130 trillion, or 15 percent of total private debt,12 with two-thirds of those loans secured by land.13 Inflated real estate values increased the risk to Japan’s unwitting lenders, as they used real estate as collateral for yet more loans—not just for CRE and mortgage loans but also for small- to medium-sized business loans and nonmortgage household loans.
As if this feedback spiral of private lending and valuation wasn’t dramatic enough, banks in this period could also own stocks. Lending brought the economic surge, which boosted the overall stock market, and that significantly increased the value of the stocks the banks held. This gave an earnings boost that augmented their lending capacity, which then accelerated private lending.14 It was a deceptively profitable equation—until the bottom fell out of stocks.
The jusen were among the most aggressive lenders during this period. Jusen were nonbank institutions formed in the 1970s by consortia of banks to make household mortgage loans since banks had mortgage limitations.15 During the 1980s, these jusen had gone beyond their household mortgage mandate to begin making CRE loans and had done so as aggressively as any other lenders.16
Because banks could not legally directly own controlling shares in other financial institutions, the jusen were not technically their subsidiaries. Smaller financial institutions, such as credit cooperatives, invested in the jusen, while the larger depository firms lent to the jusen. Too many of the larger institutions that financed the jusen steered them toward loans to risky real estate projects.17 While the larger financial companies had to adhere to regulations, the jusen operated largely in an unfettered market.18
A close insiders’ network of lenders, business executives, politicians, and bureaucrats fostered Japan’s aggressive growth. Japanese securities firms, including its largest, Nomura, would reimburse losses for favored clients—with or without that client’s knowledge—because they were offering customers guaranteed investment returns. This practice would not likely to be considered except in the boom-time fever.19
The yakuza, or crime syndicates, were also in the mix. The Nomura and Nikko securities firms financed the activities of Susumu Ishii, head of the crime syndicate Inagawa-kai, when he tried to corner the shares of the Tokyu Corporation in 1989. These associations were often forged through extortion, and the loans in question were often not repaid, as noted in the 1996 article “Yakuza Settle Bad Debts with a Bullet as Japan Bubble Bursts,” from the Independent of London. In the “morality tale about Japan’s bubble economy,” the article notes, “it soon became clear” to banks when they tried, finally, to foreclose on problem loans “that they had not only been rash in the scale of their lending, but foolish in their choice of customers. . . . The yakuza have an advantage over other debtors: everyone is terrified of them.”20
There were large loans collateralized by fraudulent certificates of deposit, sometimes issued in collusion with bank staff. A customer named Keiko Fujinori notably used forged certificates of deposit to fraudulently obtain hundreds of billions of yen in loans.21 Property companies forged certificates in collusion with banking officials in an attempt to circumvent lending limitations. And even without collusion, lenders neglected due diligence on fraudulent collateral because of management pressure to meet extremely aggressive loan growth targets.
The financial tipping point was fast approaching: by 1990, total private loans had been growing significantly faster than GDP. Five-year private debt-to-GDP growth was now over 25 percent, and the overall level of private debt to GDP now exceeded 180 percent. As land prices skyrocketed to unsustainable valuations, the land that sat under Japa
n’s Imperial Palace was reputed to be worth as much as the entire state of California. Even if not entirely true, it was a plausible analysis that underscored the preposterous overvaluations and frenzied lending of the time.
Figure 3.2. Tokyo Stock Price Index (TOPIX) and Urban Land Value Index, 1970–2010 (1980=100)
Then came the correction. First, in 1990, the Nikkei stock market average dropped by an alarming 39 percent. As in the Great Depression, the stock market crash was a symptom rather than a cause. Behind the story of both market crashes lurked the prequel of overlending and private debt. Stocks had reached unsustainable valuations and a critical mass of bad loans had long since been made. As is often true with financial crises, the rise in the stock market had accompanied the lending boom. Likewise, the stock market’s tumble was at first a symptom of peril and a harbinger of the real estate tumble, and then became a catalyst for more financial distress.
The bottom fell out of the real estate market in 1991. Whether office buildings, apartments, hotels, houses, condominiums, or tracts of land, from early 1991 to March of the next year, real estate price declined on average 15.5 percent. The trend continued into 1993, with values falling by another 18 percent by March 1993. Falling real estate prices are often a defining moment that comes pursuant to a lending surge. Prices had risen by some 300 percent, which meant that the return to preboom values would be every bit as dramatic.
The problem of course was that banks had used real estate as collateral based on very high valuations, so the moment those valuations began to decline, the loans were no longer fully collateralized. And the more real estate prices plunged, the bigger the banks’ bad debt exposure. The only way to maintain prices was to continue to make real estate loans using those high valuations. It’s the crisis trap. Continued lending can prop up real estate values, but as that real estate lending slackens, values decline and impair those loans.
At this point (and probably as early as 1990), a financial crisis in Japan was inevitable. Over ¥340 trillion of loans and debt securities had been extended from 1985 to 1990, or roughly 2.4 trillion of that era’s dollars. Time would show that well over ¥90 trillion of these loans would be bad. The banking industry only had ¥35 trillion in capital and reserves. A huge number of buildings and houses sat empty, and the day of reckoning for those loans would come. The government could not prevent the calamity. It could only influence how rapidly the bad news would unfold and how large the eventual damage would be.
With the real estate crash, Japan’s economy was in crisis and had reached a crossroads. In financial crises, expansion plotlines are alike, while bust plotlines vary. It’s in this second phase that Japan diverged most dramatically from the Great Depression.
In the United States in the early 1930s, the government stood aside and provided almost no support as bank deposits and loans contracted, and so banks failed, bringing down both good and bad businesses in the process. GDP contracted by almost 50 percent, unemployment soared, and cities were filled with Hoovervilles and bread lines. Private debt declined by a staggering 20 percent in four years. Some officials believed that this contraction was a necessary purge.
In the early 1990s, Japan’s economy could have taken that same path. It had created as much or more bad debt than in the United States in the 1920s. The level of unemployment, the damage to its banks, and the contraction in its GDP could have been every bit as severe. At this point many of the banks could have and perhaps should have been declared undercapitalized or insolvent. But instead, in stark contrast, bankers and their regulators appeared to look the other way, and Japan’s loans and GDP continued to expand, albeit at a slower pace. As part of this, banks took loan losses against their earnings that were astonishingly smaller than their actual problem loans because acknowledging the full extent of the problem would surely have meant the failure of many of the largest institutions.
Japan’s banking industry and government remained in denial longer and more deeply than those of the United States in the 1930s. Japan was all too willing to believe that property values would rebound and robustness would return. There were no “runs” since that posture and the government’s implicit support kept the industry’s funding intact, and the government also boosted deposit insurance strategically through this period. Though the collapse in asset prices occurred in 1990 and 1991, Japan’s government did not intervene to provide capital to the industry until 1998 and took until the mid-2000s to complete the process—a period of almost fifteen years.
There were several factors behind this denial. For one thing, the sheer speed and scale of the asset price decline was far beyond anyone’s ability to grasp—or, even if they could grasp it, to accept. The scale of the bad lending that brings a financial crisis is almost always too overwhelming to readily absorb. If anyone in Japan’s banking or regulatory community had said that the banking industry’s losses over the next eight years would be over ¥65 trillion, it would have been dismissed as lunacy, even though any clear-eyed calculations based on the actual property-value decline would have pointed to a loss of that magnitude.
Most participants could not accept that land prices would not recover and favorable growth would not return. Before the collapse of 1991, there had not been a decline in real estate values since World War II, much land was closely held and yielded limited information on sales prices, and the Japanese broadly believed values would never decline. And in a climate where the Japanese expected and required themselves to continue their record of superior achievement, peer pressure spurred them to keep growing. No bank wanted to be the first to show a decline in loans or earnings, and banks learned to postpone the recognition of loan losses through too-optimistic projections and other obfuscations.22
Remarkably, banks in Japan at this time had nearly complete discretion in how much they reserved for loan losses and for charging those loans off against earnings, and were not required to disclose problem loans. They were allowed to report income from a loan for a year after they stopped receiving interest payments on that loan. Further, since banks were the biggest taxpayers in Japan, they had to ask permission to reserve for loan losses for each loan because to do so would reduce their tax payments to the government. Banks actually reduced loan loss reserves in fiscal year 1991.23
But even this deep level of denial had its limits. Some problems proved impossible to suppress. In 1991, Toho Sogo Bank, despite being covered by the government’s deposit insurance system, went bankrupt and became Japan’s first bank failure since World War II.24 The government treated this small bank failure as an aberration.
The finance ministry had introduced a new rule in 1990 to try to moderate what was tacitly recognized as rampant CRE overlending. It required that banks not increase their lending to property companies by more than the percentage increase in their total loans.25 Beginning in 1991, this contributed to an overall deceleration of CRE lending but brought a shift in lending to other categories, especially household mortgages, which then became the fastest-growing category of lending. This in turn brought overlending and credit problems in the household mortgage sector. And in fact, from 1995 to 1998, when the industry was finally recapitalized, overlending in consumer mortgages and housing loans approached the overlending that had occurred in CRE.
In 1995, Japan finally began to face up to some household mortgage and CRE issues when it intervened to address the jusen problem. By 1995, nonperforming loans for the jusen had ballooned to 75 percent of all loans outstanding, with 60 percent marked unrecoverable, according to a Ministry of Finance investigation: the findings were that the situation was so bad that the ministry liquidated the jusen.26 Parent firms and creditor banks of jusen companies bore the brunt of these losses, writing off all their equity stakes and loans to the companies totaling ¥3.5 trillion.27 Other banks wrote off ¥1.7 trillion in loans.28 Ultimately, this cost taxpayers an estimated ¥1.2 trillion.29
But total loans outstanding at the jusen were below 2 percent of all private loans, so the
threat from the jusen alone was well short of systemic. The public pointedly decried the use of taxpayer dollars to rescue the jusen, which made the government reluctant to rescue any other lenders over the next two years and fed the bias toward denial. So banks, though chock-full of problem loans, kept on lending.30
By July 1995, public concern about the condition of all Japan’s banks was growing.31 To quiet these concerns, the Ministry of Finance announced that Japan’s provider of deposit insurance, the Deposit Insurance Corporation, would protect all deposits for five years, replacing the ¥10 million limit that had previously been in place. In 1996, with concerns unabated, it was further amended to protect all deposits and other liabilities—such as bonds—until March 2001. Meanwhile, the business environment continued to be difficult. The 1980s’ overbuilding had not been absorbed and in some areas had worsened. The high debt load accumulated in the 1980s and early 1990s stifled businesses and households.
The household lending upturn of the 1980s had resulted in an increase in personal bankruptcies in the early 1990s. Bankruptcies in Japan reached ¥8 trillion in 1991.32 A growing number of companies began restructuring their debt. Even in the face of this, however, no banks made significant increases to their loan loss reserves. Since selling real estate caused property values to fall even further, and falling property values impaired the loan collateral, banks encouraged debtors not to sell real estate serving as collateral since it would only add to this problem. They simply extended their loans instead.33 It’s a strategy that is sometimes pejoratively referred to as “extend and pretend.” Because of this, bank loans didn’t contract in the early to mid-1990s, and therefore the majority of businesses had the lending support to continue to muddle through. Unemployment stayed low.
The term used when a government and its regulatory departments show leniency and allow banks to forgo or take extended time to write down impaired loans is forbearance, and Japan showed extraordinary levels of this forbearance toward the nation’s lenders in the 1990s. And so these “zombie” banks, as they were now sometimes called, marched on.34
A Brief History of Doom Page 10